Bug 723536

Summary:	SEGV occurences when something fails during --remote usage on i386,ppc,s390x
Product:	Red Hat Enterprise Linux 6	Reporter:	Petr Muller <pmuller>
Component:	systemtap	Assignee:	Josh Stone <jistone>
Status:	CLOSED ERRATA	QA Contact:	qe-baseos-tools-bugs
Severity:	medium	Docs Contact:
Priority:	medium
Version:	6.1	CC:	fche, jistone, mjw, ohudlick, scox, wcohen
Target Milestone:	rc
Target Release:	6.2
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	systemtap-1.6-1.el6	Doc Type:	Bug Fix
Doc Text:	Cause The --remote did not have a proper copy-constructor for systemtap_session. Consequence SystemTap stap command would produce a SEGV andfail to build instrumentation when the --remote options was used for i686, ppc, and s390x. Fix Systemtap fixed the session code to properly copy kernel/arch information to avoid using implicit behavior for the --remote option. Result SystemTap stap command builds instrumentation when the --remote options is for i686, ppc, and s390x.	Story Points:	---
Clone Of:		Environment:
Last Closed:	2011-12-06 15:18:17 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Petr Muller 2011-07-20 13:56:41 UTC

Description of problem:
When using --remote, some stap failures cause SIGSEGV on several architectures: i386, ppc, s390x. The easiest reproducer seems to be usage of nonexistent file:

# stap --remote root@$SERVERS -v foobar
Input file 'foobar' is empty or missing.
Pass 1: parsed user script and 73 library script(s) using 22048virt/13280res/2380shr kb, in 230usr/10sys/245real ms.
Pass 1: parse failed.  Try again with another '--vp 1' option.
Segmentation fault (core dumped)

(SERVERS contains a legitimate --remote target: a compatible and ssh-accessible box)

Version-Release number of selected component (if applicable):
systemtap-1.4-6.el6_1.2

How reproducible:
always

Steps to Reproduce:
1. stap --remote root@$SERVERS -v foobar
2.
3.
  
Actual results:
SEGV

Expected results:
no SEGV

Additional info:
Does not happen on x86_64
Also, this seems to be fixed in current upstream systemtap. Filing this bug so we can track if it is fixed in 6.2.

Comment 1 Josh Stone 2011-07-20 23:00:18 UTC

I believe that commit abc15290 fixed this by adding a proper copy-constructor for systemtap_session.  This commit is included in systemtap 1.5.  Before that, some of the members were getting badly copied when we would clone for a target machine different than local.

For bonus points, it appears that i686 targets always got cloned, because the remote "i686" was compared to the normalized local "i386" and deemed different.  That's why x86_64 may not have this problem, because there aren't any variant arch names for x86_64.  I suspect if you had used an x86_64 target with a different kernel version than local, it would also fail on 1.4.  Anyway, commit ba0bd1be will make sure to normalize the arch name before a clone is attempted.

Comment 4 William Cohen 2011-11-11 21:58:41 UTC

    Technical note added. If any revisions are required, please edit the "Technical Notes" field
    accordingly. All revisions will be proofread by the Engineering Content Services team.
    
    New Contents:
Cause
    The --remote did not have a proper copy-constructor for systemtap_session.
Consequence
    SystemTap stap command would produce a SEGV andfail to build
    instrumentation when the --remote options was used for i686, ppc, and s390x.
Fix
    Systemtap fixed the session code to properly copy kernel/arch
    information to avoid using implicit behavior for the --remote option.
Result
    SystemTap stap command builds instrumentation
    when the --remote options is for i686, ppc, and s390x.

Comment 5 errata-xmlrpc 2011-12-06 15:18:17 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHBA-2011-1517.html